mirror of
https://github.com/tcsenpai/agenticSeek.git
synced 2025-06-07 03:25:32 +00:00
57 lines
1.6 KiB
Markdown
57 lines
1.6 KiB
Markdown
# localJarvis
|
|
|
|
A fully local assistant using swarm of deepseek agents, with multiple capabilities such as code execution, web browsing, etc...
|
|
|
|
THIS IS A WORK IN PROGRESS
|
|
|
|
## Install
|
|
|
|
- Make sure you have ollama installed on your machine
|
|
- Install dependencies (`pip3 install -r requirements.txt`)
|
|
|
|
## Run fully local
|
|
|
|
Simplest way is to use ollama
|
|
- First change the config.ini file to set the provider_name to `ollama` and provider_model to `deepseek-r1:7b`
|
|
- In first terminal run `ollama serve`
|
|
- In second terminal run `python3 main.py`
|
|
- Ollama will download `deepseek-r1:7b` on your machine
|
|
- 2 model are also downloaded:
|
|
* For text to speech: `kokoro`
|
|
* For speech to text: `distil-whisper/distil-medium.en`
|
|
- type or say goodbye to exit.
|
|
|
|
# Run model on another machine
|
|
|
|
- First change the config.ini file to set the provider_name to `server` and provider_model to `deepseek-r1:7b` (or higher)
|
|
- On the machine that will run the model execute the script in stream_llm.py
|
|
|
|
```
|
|
python3 stream_llm.py
|
|
```
|
|
|
|
- In the config.ini file, set the provider_server_address to the ip address of the machine that will run the model.
|
|
|
|
- On the machine that will run the assistant execute main.py
|
|
|
|
```
|
|
python3 main.py
|
|
```
|
|
|
|
## Text to speech
|
|
|
|
If you want your AI to speak, run with the `--speak` option.
|
|
|
|
```
|
|
python3 main.py --speak
|
|
```
|
|
|
|
## Current capabilities
|
|
|
|
- All running locally
|
|
- Reasoning with deepseek R1
|
|
- Python code execution capabilities
|
|
- Bash execution capabilities
|
|
- Get feedback from python/bash interpreter attempt to fix code by itself.
|
|
- Fast text-to-speech using kokoro.
|